Basel III fragments as US, Europe and India take diverging paths on bank capital rules

When Basel 3 accord was finalised in 2010, its ambition was to turn global banking into a system governed by common rules rather than national quirks. Fifteen years later, that orderly garden looks more like a jungle: dense, overgrown and navigable only with local knowledge.

Basel 3 targeted the 'black box' of internal models, where identical assets produced divergent capital requirements. Regulators responded by constraining models, standardising inputs and adding a backstop - the output floor - ensuring model-based capital stayed above 72.5% of standardised levels. The goal: a common language for risk. That language is now fragmenting.

The latest chapter came from the US on March 19. After years of wrangling, the Fed, Federal Deposit Insurance Corporation (FDIC) and Officer of the Comptroller of the Currency (OCC) have retreated from their July 2023 proposal, which implied roughly a 19% increase in capital for the largest banks. The new approach is expected to be broadly 'capital neutral'. This reflects a shift in philosophy: from maximising resilience to balancing resilience against credit creation and competitiveness.


US regulators are concerned that pushing capital requirements too high will drive lending out of the regulated banking system and into less-supervised shadow institutions. The new framework, thus, leans toward protecting traditional bank activities, even at the cost of deviating from the Basel script.

Operational risk is where Basel 3's comparability breaks down. The framework ties capital to past losses and scale, anchoring risk in experience. The US pares this back - dropping the internal loss multiplier and allowing broad fee netting - weakening the link between misconduct and capital. The result: billions in relief for large banks, but a widening gap with European peers still bound by stricter rules.

Systemic risk shows the same pattern. Instead of stacking Basel 3 on existing surcharges, US regulators adjusted the surcharge itself - partly indexing it to growth to avoid 'double counting'. This blunts Basel's tightening, while European and British banks, lacking such offsets, face structurally higher capital burdens.

Even Basel's centrepiece, the output floor, has been sidestepped. As Europe and Britain struggle to phase in the 72.5% constraint, the US has shifted to a more standardised approach that largely bypasses it. The result is simpler rules, but diminished comparability with European banks' risk-weighted assets. What emerges is not outright deregulation but selective implementation. Basel has become less a rulebook than a menu.

In India, RBI has taken the opposite approach - treating Basel as a floor rather than a ceiling. Capital requirements are routinely 'gold-plated', with banks operating well above minimum thresholds.

Where the US seeks flexibility, India prefers discipline. Following the collapse of Silicon Valley Bank, RBI moved to strengthen liquidity rules, including higher assumed run-off rates for digitally enabled deposits. It has also ventured into new territory, introducing climate-adjusted risk weights that penalise exposure to carbon-intensive sectors.

The most consequential shift is yet to come. From 2027, India will move fully to an expected credit loss (ECL) framework, replacing the older incurred-loss model. Unlike in the US or Europe, where such models are heavily reliant on banks' internal estimates, RBI has imposed prudential floors - minimum provisioning requirements that cannot be gamed by optimistic assumptions. The result will be a system that is more conservative, more capital-intensive and less susceptible to model manipulation.

By the end of the decade, three distinct versions of Basel 3 will coexist:

US Lite American variant will be capital-neutral, market-friendly and designed to beat non-banks.

EU/Britain lag European and British versions will be strict on paper, but bogged down by endless 'transitional arrangements' and 'offsets'.

Indian fortress This will be high capital, high provisioning and zero trust in internal bank models.

The consequences are significant. Basel 3 was intended to make capital ratios comparable across borders, to allow investors, regulators and counterparties to assess risk on a like-for-like basis. That goal is now receding. As national discretion grows, capital ratios increasingly reflect local policy choices as much as underlying risk.

This may have been inevitable. Banking regulation sits at the intersection of financial stability, economic growth and political sovereignty. No country willingly outsources such decisions to an international committee. Yet, the cost of this autonomy is fragmentation. A system designed to impose discipline on global banks is becoming one in which outcomes depend as much on geography as on risk.

Basel 3 was meant to standardise the measurement of safety. Instead, it has standardised the negotiation over it. The next crisis, when it comes, will not resolve these differences gently. It rarely does.

The writer is former ED, Nomura India